FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1956
1956 - 1483.PDF
FLIGHT, 19 October 1956 637 The Armstrong Siddeley Screamer ROCKET-MOTOR DESIGN A Paper by the Chief Engineer of Armstrong Siddeley Motors Rocket Division) LAST week a Royal Aeronautical Society Main Lecture wasread at the R.A.F. Technical College before the Henlow'branch of the Society. The author was Mr. Sidney Allen, F.R.Ae.S., chief engineer (rocket division) of Armstrong SiddeleyMotors, Ltd., whose operations are based at Ansty, near Coventry. As Mr. Allen pointed out in his opening remarks, little haspreviously been published on actual experience in designing and developing a rocket motor for a given application within a giventime interval. He was therefore remedying this position by pro- viding a full and authoritative account of the problems posed inthe evolution of the Armstrong Siddeley Screamer. This is a single-chamber motor, running on liquid oxygen and aviationturbine fuel, intended for piloted aircraft application and with a design thrust controllable from 8,000 down to 800 lb. (It wasfully described and illustrated with a cutaway drawing in Flight of July 27, 1956.) In its design (said Mr. Allen) there were threemain sources of trouble: the wide thrust range required; the design of a turbo-pump unit operating on the rocket propellants;and dealing with the starting problem and at the same time main- taining the standard of safety and reliability required from thepowerplant of a piloted aircraft. It was early believed, continued the lecturer, that the main lossin specific impulse due to throttling (at high altitude) would be caused by combustion inefficiency. Nevertheless, gas-turbinedesigners were obtaining efficiencies well above 90 per cent on only a few pounds per square inch combustion pressure and itwas accordingly decided to attempt to meet the whole range of thrust required with a single chamber. In the design of theturbo-pump, high-test peroxide was rejected as a working fluid, in spite of its attractions, since its inclusion would have meant thenecessity of supplying three "strategic" liquids for the single powerplant. It was eventually decided to cut down developmenttime by injecting water (i.e., steam) to provide a suitable working fluid for the turbine. The working fluid was prepared in a gas-generator unit comprising a small combustion chamber^ with appropriate injectors and igniter assembly. In the_ final design ofgas generator it was found possible to use conventional swirl-type Portions of three of the early types of turbine studied dur- ing the development of the Screamer (in chronological order from left to right). fuel injectors, three of which were inserted through the water ductat the upstream end of the combustion zone. The liquid oxygen was swirled in around a cup containing the igniter so that theoxygen was vaporized by the igniter flame before the fuel was injected into it. The feed to the turbine had a final temperatureof about 920 deg K. An accompanying sketch shows three of the early designs ofturbine rotor which were examined. The first design had an efficiency of only 38 per cent, but this was raised to around 50 percent by increasing the number of buckets from 13 to 33 and providing a "roof" oyer each bucket to prevent the working fluidfrom spilling out radially. Eventually, however, it was decided to design a conventional axial-flow, bladed turbine operating at amean blade speed of l,100ft/sec. The final design had blades of Nimonic 80A, fed from a single supply nozzle, and showed itselfto be capable of providing adequate power with an efficiency of 60 per cent. The three propellant pumps were of similar design, being open-impeller, centrifugal units with booster screws, similar to the smaller patterns developed for the earlier Armstrong SiddeleySnarler [fully described in Flight of August 6, 1954]. Develop- ment of the liquid-oxygen pump was hampered by cavitation inthe light-alloy diffuser casing, a difficulty eventually overcome by a change to stainless steel. Another problem was that, although inthe earlier Snarler engine the bearing nearest to the liquid-oxygen impeller was run dry, its life was considered inadequate for theScreamer. Accordingly, it was decided to lubricate all the bearings with oil mist from the gearbox, the oil/liquid-oxygen seal being anoptically flat, stainless-steel ring mounted on a bellows and spring- loaded on to a sintered bronze ring impregnated with Fluon.In the design pf the combustion chamber, provision had to be made for a maximum fuel flow of about 4,000 gal/hr and it wasnot found possible to provide space for an adequate number of swirl-type sprayers. The best compromise appeared to be theprovision of plain nozzles so disposed as to cause the fuel and oxygen jets to impinge on each other. Initially it was decided toincrease the resistance of the inner shell to buckling by employing helical passage walls—provided to control the flow of regenerativecooling kerosine—as stress-carrying members machined integral with the inner shell. Development running showed that someform of film cooling would be necessary, but the scrolled inner shell did not lend itself well to such an arrangement. It was finallydecided that, as water was already being carried for the gas generator, it could also be used to cool the chamber jacket.Calculations indicated that it should be possible to cool the Screamer with water forming 13 per cent of the total propellantsand giving a theoretical specific impulse only a few points lower than that obtainable from the best mixture of oxygen and kerosine.Eventually the best design of chamber was found to be a throatless pattern with like-on-like impingement injectors. Very extensive development was also necessary to produce theright design of valves, igniters and control system in order to provide safe and reliable starting and variable-thrust running withsingle-lever pilot control. During the starting cycle the initial acceleration of the turbine up to operating speed was achieved bysupplying the liquid oxygen, kerosine and water from small pressurized tanks, each of which was designed to vent and refillautomatically after each start. In the case of the starting tanks for water and fuel, expulsion bags proved to be a simple solution, butfor the liquid oxygen it was found necessary to make the tank as a long cylinder fitted with a piston provided with Duaflex rings.The Screamer then became self-sustaining by feeding the turbine
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
